Download NAFTA at 20: Misleading Charges and Positive

NUMBER PB14-13
MAY 2014
NAFTA at 20: Misleading
Charges and Positive
Achievements
G a r y C lyd e Huf b au er, Cathleen Cimino, and Tyle r M oran
Gary Clyde Hufbauer is the Reginald Jones Senior Fellow at the Peterson
Institute for International Economics. He has written numerous books
on international trade, investment, and tax issues, including NAFTA
Revisited: Achievements and Challenges (2005). Cathleen Cimino
and Tyler Moran are research analysts at the Institute.
Authors’ note: The authors thank C. Fred Bergsten, William R. Cline,
Caroline Freund, Theodore Moran, Jeffrey J. Schott, Edwin M. Truman,
and Steve Weisman for helpful comments.
© Peterson Institute for International Economics. All rights reserved.
In truth the claims on both sides of the NAFTA issue 20
years ago were overblown. Since the Mexican economy is less
than one-tenth the size of the US economy, it is not plausible
that trade integration could dramatically shape the giant US
economy, even though integration could exert a substantial
impact on the relatively small Mexican economy. But exaggeration and sound bites are the weapons of political battle, and
trade agreements have been on the front line for two decades.
President Bill Clinton, for example, declared that NAFTA
would “create” 200,000 American jobs in its first two years
and a million jobs in its first five years. Not to be outdone,
NAFTA opponents Ross Perot and Pat Choate projected job
losses of 5.9 million, driven by what Perot derided as a “giant
sucking sound” emanating from Mexico that would swallow
American jobs.2 Both of these claims turned out to be overblown, especially the one advanced by Perot and Choate.
One reason NAFTA remains controversial
is that, for the United States, NAFTA
marked the first major trade deal with
Twenty years after its enactment, the North American Free
Trade Agreement (NAFTA) continues to divide Americans and
cast a shadow over the US trade agenda. Opponents of the most
recent free trade agreements with Colombia and South Korea
repeatedly cited NAFTA as a malignant precedent, charging
that NAFTA cost millions of US jobs, suppressed wages, and
deepened US economic inequality, and claimed that new trade
agreements would do the same. Today NAFTA is being invoked
again in debates over Trade Promotion Authority (TPA), which
would give President Obama latitude to negotiate new trade
deals, specifically the Trans-Pacific Partnership (TPP) and the
Transatlantic Trade and Investment Partnership (TTIP) (see
Scott 2013 and 2014). One critic of the TPP recently labeled it
“NAFTA on steroids.”1
1. Lori Wallach, “NAFTA on Steroids,” Nation, June 27, 2012, http://www.
thenation.com/article/168627/nafta-steroids# (accessed on March 13, 2014).
1750 Massachusetts Avenue, NW
Washington, DC 20036
a poor country, namely Mexico.
In recent debates, NAFTA seems to have had few vocal
defenders. Yet because of the central role it continues to play
in the US consciousness, this is an opportune moment to
separate fact from fiction in the long-running disagreement
over NAFTA. The purpose of this Policy Brief is not to rehash
old claims that may have been overstated but to clear the air
so that the benefits and challenges of trade can be examined
in an objective light.
NAFTA took effect on January 1, 1994, alongside the
previously negotiated Canada-US Free Trade Agreement
(CUSFTA). NAFTA committed the United States and
Mexico to eliminate all US and Mexican tariffs over a tenyear period, except on a handful of agricultural exports that
2. The contemporaneous debate is summarized in Hufbauer and Schott
(2005).
Tel 202.328.9000
Fax 202.659.3225
www.piie.com
NUMBER PB14-13
were to be phased out over 15 years. The accord also aimed
to lower cross-border barriers to services and investments
while setting standards for patents, trademarks, and other
forms of intellectual property rights. One reason that NAFTA
remains controversial is that, for the United States, which had
previously embraced a series of global trade accords after the
Second World War, NAFTA marked the first major trade deal
with a poor country, namely Mexico.
The NAFTA partners encountered rough waters in the
pact’s inaugural year, and enduring perceptions of NAFTA were
adversely shaped by three Mexican shocks. First, on January 1,
1994, the Zapatista rebellion erupted in the southern Mexican
state of Chiapas. While the rebellion had little direct connection to NAFTA provisions, it was deliberately timed with the
NAFTA played a decisive role in the
recovery of the Mexican economy by
fostering a large financial rescue package
and enabling a sharp turnaround in
Mexico’s external trade balance.
pact’s entry. One of the rebels’ many grievances was opposition to NAFTA for providing a “symbolic manifestation of the
huge attention the Mexican government paid to the modern
northern states and the neglect of the historically poor southern
states” (Hufbauer and Schott 2005 p. 10). On March 23, 1994,
the Chiapas uprising was followed by the assassination of presidential candidate Luis Donaldo Colosio, the heir apparent to
then Mexican President Carlos Salinas. The culprits were never
identified, and the assassination triggered alarms in the investor
community. Finally and most damaging came the abrupt and
progressively more severe devaluation of the Mexican peso,
initially on December 20, 1994. The peso crisis followed the huge
buildup of debt, denominated in US dollars, issued both by the
Mexican government and Mexican firms, to finance a widening
current account deficit. Beginning in the spring of 1994, investors began fleeing Mexico, depleting the Banco de Mexico’s
holdings of foreign exchange, as the central bank attempted to
defend the peso’s fixed rate to the dollar. When the peg was
finally abandoned in December 1994, the steep devaluation of
the peso led to a collapse of imports and a surge of Mexican
exports. For many NAFTA critics, the “temporal connection
between NAFTA ratification and Mexico’s economic collapse
was too powerful to be mere coincidence” (Hufbauer and
Schott 2005 p. 9). While bad policy choices in preceding years
had set the stage for the financial crisis, NAFTA was blamed
2
MAY 2014
for inadequate monitoring of Mexico’s macroeconomic policies. While this criticism had some merit, NAFTA also played
a decisive role in the recovery of the Mexican economy, both
by fostering a large financial rescue package and by enabling a
sharp turnaround in Mexico’s external trade balance. 3
Despite NAFTA’s inauspicious launch and subsequent
charges made against it, the agreement can be credited with
making important strides toward intraregional integration and
higher living standards in all three countries. The interdependence of the United States, Canada, and Mexico is striking. For
example, goods imported from Canada are estimated to contain
25 percent of US inputs and from Mexico, 40 percent of US
inputs (Koopman, Powers, Wang, and Wei 2010). In 2013,
about 14 percent of US merchandise exports went to Mexico,
exceeding the combined total of merchandise exports to
Germany, France, the United Kingdom, and the Netherlands.
Since 1993, US trade with Mexico quintupled in nominal
terms, whereas trade with the rest of the world increased three
times. NAFTA promoted the integration of the regional energy
market—particularly between the United States and Canada—
which somewhat mitigated US reliance on imports from sources
across the Atlantic, while encouraging greater energy independence within the region.4
Many US jobs depend on exports—an estimated 2.6
million on exports to Canada and 1.9 million on exports to
Mexico.5 Following the approval of NAFTA, Mexico went
into a financial crisis that discredited its policies in the eyes
of many. But the mid and late 1990s were a period of boom
times in the United States, and fears that NAFTA would
cause a surge of unemployment subsided. Indeed almost 17
million jobs were added to the US economy in the seven years
following enactment of NAFTA, and the unemployment rate
dropped from 6.9 percent to 4.0 percent.
On the other hand, the last two decades have seen growing
inequality in the United States and concerns that low-skilled
jobs have been hollowed out both by advances in technology
and the signing of trade agreements. Inevitably, in the 2000s,
3. For a longer discussion, see Hufbauer and Schott (2005), pp. 8–12.
4. The US shale revolution could well convert the United States into a net
exporter in the coming decade, thereby altering the traditional dynamic of
intraregional energy trade. Partial privatization of oil and gas production in
Mexico might also release substantial new supplies. The combination of shale
energy and Mexican liberalization could rapidly bolster the prospects of physical energy independence in North America.
5. The job estimates assume a coefficient of 7,500 direct and indirect jobs per
billion dollars of exports. US exports to Canada in 2013 were $366 billion,
and US exports to Mexico were $256 billion. The jobs coefficient is derived
from the input-output analysis reported in table 2 of Lawrence (forthcoming
2014). Direct export jobs are approximately 5,100 per billion dollars of
exports, and indirect jobs are another 2,400.
NUMBER PB14-13
NAFTA again became a proxy for fear over job losses. But
concerns about jobs during the initial NAFTA debate were
badly distorted, and misstatements then are repeated today.
It is widely understood that an expansion of two-way trade
will shuffle jobs between sectors of the economy: Importcompeting sectors will lose some jobs and export-oriented
sectors will gain some. Yet most economists took the view
that the net number of jobs gained or lost owing to NAFTA
would be statistically insignificant in a US labor force that
then numbered 110 million. In their analysis, Hufbauer and
Schott (1993) calculated that the agreement could create
170,000 net US jobs “in the foreseeable future.” Advocates
of NAFTA, including those at the Peterson Institute for
International Economics, argued that the main payoff from
NAFTA would be better jobs, not more jobs, as the US and
Mexican economies were restructured according to the law of
comparative advantage. But what economists had to say was
lost in the political din of the 1990s and is often ignored in the
contemporary debate over TPP and TTIP.
Economic analysis of the channels by which trade agreements potentially lead to higher national output has made
significant advances since the 1990s,6 as has the understanding of the costs of job churn that inevitably accompanies
economic restructuring in the wake of trade liberalization.7
(Churn refers to the phenomenon of large numbers of workers
loosing and gaining jobs over a fixed time period.) Yet the
US political rhetoric surrounding trade agreements essentially
channels the NAFTA debate of two decades ago. Proponents
claim job gains and higher living standards; opponents claim
job losses, lower wages, and corporate enrichment.
No proponent argues that North America entered a
golden age after NAFTA. But critics are wrong when they
blame NAFTA for ills that should not be laid at the agreement’s doorstep and wrong when they dismiss the genuine
achievements of the tripartite pact.8 In this Policy Brief, we
first answer six charges voiced by NAFTA critics and then
sketch the positive case. The six central charges and our short
responses are:
 NAFTA fostered a growing US trade deficit.
 Short response: Not perceptible.
 Trade with Mexico raised US unemployment.
 Short response: Not perceptible.
6. For a survey of the multiple payoff channels to the United States, see
Bradford, Grieco, and Hufbauer (2005). For a summary of the payoff channels
from foreign direct investment to a developing country host economy, see
Moran (2011).
7. Lawrence (forthcoming 2014) summarizes the literature on job churn.
8. For examples of ills laid at NAFTA’s doorstep, see Public Citizen’s Global
Trade Watch (2014).
MAY 2014
 Job loss depressed US wages, especially in manufacturing.
 Short response: In some cases, but not across the
board.
 The boom in US agricultural exports turned rural Mexicans
into illegal emigrants.
 Short response: No connection.
 Apart from agriculture, NAFTA abetted illegal immigration.
 Short response: The opposite.
 Mexican growth has not achieved the rate anticipated by
NAFTA proponents.
 Short response: A fair criticism.
While dubious at best, these charges have been repeated
so often that they have congealed into conventional wisdom
and are parroted even by mainstream journalists.9 Before
addressing the charges, it’s worth emphasizing that they are
all directed at the US-Mexican experience. Yet NAFTA is a
tripartite pact, and hardly anyone criticizes the US-Canada
experience. In fact two-way trade and investment outcomes
across the northern US border have been strong and almost
uniformly positive.
U S T R A D E D E F I C I T W I T H M E X I CO
Larger US trade deficits are often cited by critics of NAFTA
and other trade agreements as a sure consequence of these
pacts and an unhealthy outcome for the United States. Political
leaders frequently decry trade deficits, arguing that exports
support jobs at home whereas imports substitute for products
that could be produced by American workers. These broadside
attacks against trade deficits are misguided. Bilateral US trade
deficits are not necessarily bad. In a world of multilateral trade,
even if the United States achieved overall balance in its external
accounts, US trade would not be in balance with each country.
For example, the United States might be in deficit with Mexico
but in surplus with Canada.
Those who measure the “success” of preferential trade
negotiations in terms of the consequent bilateral trade balances
among the participants overlook the fact that it is logically
impossible for all members of a preferential trade agreement to
end up with bilateral trade surpluses as a result. This observation pertains just as much to the upcoming TPP and TTIP
negotiations as to NAFTA. It is possible that some members
of a preferential trade agreement will improve their global trade
9. See, for example, Christopher Caldwell, “Popular sentiment is hardening
against free trade,” Financial Times, February 28, 2014, http://www.ft.com/
intl/cms/s/0/a538be0a-9ee9-11e3-8663-00144feab7de.html# axzz2vJckoimC
(accessed on March 3, 2014).
3
NUMBER PB14-13
Figure 1
MAY 2014
US-Mexico bilateral merchandise trade (nonfuel), 1993–2013
billions of current US dollars
300
Mexican exports to the
United States
US exports to Mexico
250
200
150
100
50
0
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
Note: Mexican exports to the United States are based on US import data. Nonfuel merchandise trade based on Standard International Trade
Classification (SITC) codes minus SITC 3 (mineral fuels/lubricants).
Source: World Bank’s World Integrated Trade Solutions (WITS) database, https://wits.worldbank.org/.
balances after the completion of an agreement, but this will be
because of two distinctive factors: first, increased efficiency in
the use of resources and second, internal changes in the relationship between income and expenditure within the partner
state.
As explained in the appendix to this Policy Brief, the United
States is bound to run an overall trade deficit with the rest of the
world when combined US savings of the household, business,
and government sectors are negative, as they have been for some
years. To finance the trade deficit, the United States is obliged
to borrow from the rest of the world. In such circumstances, a
global US trade deficit is inevitable. At best, trade agreements
exert a second-order impact, possibly changing the pattern of
bilateral surpluses and deficits but exerting a marginal impact
on the size of the global trade deficit.
With these precepts in mind, a look at US two-way trade,
and the trade deficit with Mexico since NAFTA was agreed
in 1993, may be instructive. Our analysis excludes petroleum
and natural gas trade from the picture for the simple but
powerful reason that if the United States did not import petroleum and gas from Mexico and Canada, it would import fuels
at higher cost from other countries. Figure 1 charts US exports
of goods and services to Mexico, and Mexican exports to the
United States, excluding fuel in both directions. Two-way
trade has expanded enormously, by a factor of five in current
4
dollars. Just as the critics say, the US bilateral trade deficit with
Mexico has also grown, going from a surplus of $5 billion
in 1994 to a deficit of $45 billion in 2013.10 But this was
not because of a “giveaway” deal by US trade negotiators. As
appendix table A.1 shows, at the time NAFTA was launched,
the average US tariff on imports from Mexico was 4.3 percent,
while the average Mexican tariff on imports from the United
States was 12.4 percent. Since both tariff averages went to zero
fairly quickly, the country “giving away,” measured by tariff
concessions, was Mexico, not the United States.
The main reason for the growing US bilateral trade
deficit with Mexico over two decades was the growing imbalance between income and spending within the United States.
Reflecting this widening imbalance, between 1994 and 2013,
the US nonpetroleum goods deficit with the world expanded
from $120 billion to $510 billion.11 The global enlargement
of the trade deficit is not an outcome of NAFTA or other
free trade agreements, as the appendix clearly shows. Rather,
10. Trade data from the World Bank’s World Integrated Trade Solutions
(WITS) database. Adjusted for inflation, two-way trade grew by a factor of
three between 1994 and 2013. The 1994 US bilateral surplus with Mexico
was largely illusory because, in that year, Mexico was running an unsustainable
current account deficit with the world.
11. The 2013 figure is based on the first three quarters of 2013 (US Bureau of
Economic Analysis).
NUMBER PB14-13
it reflects the fact that the United States gradually altered its
status from small net borrower to huge net borrower driven
largely by rising federal budget deficits and falling household
savings. Moreover, the Mexican share of the 2013 global
US goods deficit, about 9 percent ($45 billion versus $510
billion), was much smaller than the Mexican share of US
nonpetroleum merchandise imports, about 13 percent ($248
billion versus $1,939 billion). As mentioned, when NAFTA was launched, the average
Mexican tariff was much higher than the average US tariff:
12.4 percent versus 4.3 percent.12 In 1993, Mexican nonpetroleum exports to the United States were $39 billion, and US
nonpetroleum exports to Mexico were $41 billion. Applying
a simple trade-elasticity approach, and assuming an elasticity
of 3.0 (a high value), eliminating these average tariffs would
suggest an increase of $5 billion in Mexican exports to the
United States and an increase of $14 billion in US exports to
The main reason for the growing US bilateral
trade deficit with Mexico over two decades
was the growing imbalance between income
and spending within the United States.
Mexico. Such calculations made it seem likely, in the view of
some analysts, that US exports to Mexico would expand much
more than Mexican exports to the United States. Nonetheless,
the opposite happened, but for reasons that had little to do
with the warnings of critics. The main reasons were Mexico’s
newfound openness to investment, much-improved access to
US parts and components (owing to lower tariffs), and regulatory reforms.
Just as NAFTA was being implemented, in late 1994,
Mexico was decimated by the unforeseen peso crisis (devaluing the peso from 3.97 to the dollar in December 1994 to
7.76 to the dollar in December 1995). The peso crisis erupted
because the Mexican government and firms, in the preceding
two years, had imprudently issued tens of billions of debt
effectively denominated in US dollars.13 Seeing an unsustainable situation, Mexican and foreign investors alike headed for
the exits during the spring and summer of 1994, depleting
the central bank’s holdings of foreign exchange as it tried to
maintain a fixed rate between the peso and the dollar. When
12. Tariff figures from the UN Conference on Trade and Development Trade
Analysis and Information System (UNCTAD TRAINS) database.
13. Mexican government short-term debt, called tesobonos, contained an
exchange rate guarantee clause that linked the debt to the US dollar. Mexican
firms simply borrowed in dollars.
MAY 2014
the central bank abandoned the exchange rate peg, the sudden
peso devaluation led to a collapse of imports and a rise of
exports. Mexicans could no longer afford American shopping trips to San Diego and, facing a depressed home market,
Mexican firms did their best to sell into the US market. The
nonpetroleum US bilateral trade surplus of $5 billion in 1994
turned into a deficit of $12 billion in 1995.
To its credit, the Mexican government responded to the
crisis by cutting regulations that prevented foreign investors
from coming into the country, accelerating a trend that had
started earlier. Mexico laid out the welcome mat for foreign
investors of all nationalities, rather than just its NAFTA partners. Foreign companies (led by US auto firms) expanded
their plants in Mexico, integrating them with all of North
America. The outcome was a sustained burst in Mexican
exports, enlarging the nonpetroleum US bilateral trade deficit
from $12 billion in 1995 to $45 billion in 2013.
What if NAFTA had never been agreed? Would the US
trade deficit with the world be $45 billion lower in 2013? Most
unlikely. Viewing the no-NAFTA scenario from a macroeconomic perspective, a lower deficit with Mexico would have
been like squeezing a balloon—most of the deficit would
have popped out someplace else, because US expenditures
would still have exceeded US earnings by ever larger amounts
during the late 1990s and 2000s. Viewing the no-NAFTA
scenario from a microeconomic perspective, US and foreign
companies, in search of lower costs for their worldwide supply
chains, would probably have opened additional plants elsewhere in Latin America and Asia.
To conclude, it may well be true—thanks initially to the
peso crisis and over a longer period to Mexican reforms—that
NAFTA fostered a larger bilateral trade deficit with Mexico.
But it is not true that NAFTA fostered an equally larger US
trade deficit with the world.
T R A D E W I T H M E X I CO A N D U S
U N E M P LOY M E N T
With or without trade, over 4 million Americans are separated
involuntarily from their jobs each year by plant shutdowns and
mass layoffs, even when the United States is adding overall jobs
to the national payroll. But only a small fraction of the jobs
lost are caused by imports in general or imports from Mexico.
Growing US trade with Mexico (and with the world) clearly
contributes to churn in the US job market, but trade is hardly
the sole explanation. About 5 percent of this job churn (around
200,000 workers annually) can be explained by rising trade
with Mexico (discussed below). Two-way trade expands some
industries and shrinks others; this is the real-life face of compar-
5
NUMBER PB14-13
ative advantage. Within industries, growing trade downsizes
less efficient firms and upsizes more efficient firms; this is the
real-life face of the “sifting and sorting” phenomenon better
understood since NAFTA was ratified (see Bernard, Jensen, and
Schott 2003). Empirical evidence demonstrates that comparative advantage and sifting and sorting are exactly what happen
when two-way trade grows. The inevitable outcome is that some
Americans lose their jobs, identifiably because of increased
imports, while other Americans gain new or better jobs, far less
identifiably but because of increased exports.
Identifiable job losses are the stock-in-trade of NAFTA
critics. At the Economic Policy Institute, Scott (2011) estimates
that between 1994 and 2010 nearly 683,000 US jobs were lost
due to US trade deficits with Mexico (about 40,200 jobs per
year). His estimates of the net employment impact of NAFTA
use direct and indirect labor requirements of producing output in
a given industry and assume that industry trade deficits displace
At most, 5 percent of dislocated [US] workers
can be traced to imports from Mexico.
domestic production (and thus labor), dollar-for-dollar. Other
accounts are more anecdotal. In 1997, Public Citizen interviewed more than 60 US companies and found that just three
years after NAFTA, 90 percent of the promises made by proNAFTA companies to create domestic jobs or expand exports
were not fulfilled, citing General Electric, Johnson & Johnson,
Siemens, and Xerox, among others that laid off workers or
shut down facilities and shifted production to Mexico.14 Public
Citizen (2014) also points to tangible job losses based on the
number of workers receiving Trade Adjustment Assistance
(TAA), reporting that more than 845,000 workers were certified
for TAA based on jobs lost to imports from Canada and Mexico
or relocated factories between 1994 and 2013, on average about
44,500 per year.
Moreover, the general perception that job losses are associated with free trade agreements remains strong. A 2010 survey
conducted by the Pew Research Center on the public view of
FTAs, including NAFTA, found that 55 percent of respondents held the view that FTAs lead to US job losses, while
only 8 percent view trade pacts as supporting job creation (24
percent said FTAs make no difference).15 Public perceptions
14. “NAFTA’s Broken Promises: Failure to Create U.S. Jobs,” Public Citizen’s
Global Trade Watch, January 1997, http://www.citizen.org/trade/article_redirect.cfm?ID=1767 (accessed on April 2, 2014).
15. “Public Support for Increased Trade, Except with South Korea and China:
Fewer See Benefits from Free Trade Agreements,” Pew Research Center,
November 9, 2010, http://www.people-press.org/2010/11/09/public-supportfor-increased-trade-except-with-south-korea-and-china/ (accessed on April
6
MAY 2014
mirror the observation that jobs supported by exports are
often invisible and forgotten.
Job displacement is painful for the losers, but it pays off
enormously for Americans as a whole. According to calculations by Robert Z. Lawrence, looking just at US trade with
China over the last decade, for every net manufacturing
job lost to trade with China (taking into account both jobs
displaced by imports and jobs supported by exports), the US
economy gained about $900,000 in 2008. The gains reflect
enhanced productivity, a broader range of goods and services,
and lower prices at the checkout counter for households.16 The
arithmetic of national gains relative to net jobs lost would be
roughly similar for US trade with Mexico: several hundred
thousand dollars of gains to the economy for every net manufacturing job lost.
However, for individual workers facing import competition, what counts most is “jobs displaced,” not “net jobs lost.”
Between 1994 and 2013, US imports from Mexico (many
of them parts and components used in American plants)
expanded from $48 billion to $302 billion.17 In recent years
(2009 through 2013), the expansion has averaged about $27
billion annually. The direct and indirect US labor equivalent
of every billion dollars of imports is currently about 7,500
workers.18 What these numbers imply is that in recent years
additional imports from Mexico displaced about 203,000 jobs
that are lost annually to the churn. These are painful numbers
for displaced workers. However, in the overall picture of
involuntary job churn, the contribution of Mexican imports is
small. From the beginning of 2009 to the end of 2011, about
13 million workers were “dislocated” (meaning the victims
of mass layoffs), indicating an annual dislocation of about
4 million workers—mainly because of technological and
competitive forces within the giant US economy. 19 At most,
5 percent of dislocated workers can be traced to imports from
Mexico. Moreover, the churn number associated with imports
from Mexico, about 203,000 jobs displaced annually, is much
2, 2014). Similarly, 45 percent have the view that FTAs lower wages in the
United States, while 8 percent view FTAs as increasing wages.
16. The $900,000 figure is based on estimated gains to the US economy from
Chinese manufacturing imports of 0.6 percent of US GDP in 2008, which
works out to $88 billion, or about $250 per US citizen. See Lawrence (forthcoming 2014). A full explanation of the channels by which increased two-way
trade delivers gains to the US economy can be found in Bradford, Grieco, and
Hufbauer (2005).
17. Trade figures from the US Bureau of Economic Analysis.
18. The figures are derived from US manufacturing imports and their direct
and indirect US employment equivalent (total jobs) in 2012; see table 2 from
Lawrence (forthcoming 2014).
19. US Bureau of Labor Statistics, “Worker Displacement: 2009–2011,” news
release, http://www.bls.gov/news.release/pdf/disp.pdf (accessed on March 7,
2014).
NUMBER PB14-13
larger than the “lost jobs” number calculated either from trade
deficits or TAA certifications (about 45,000 jobs lost annually). The reason is that many displaced workers land another
job within a short period of time.
But focusing on jobs lost through imports is only half
the story. It is important not to forget the export side of the
job equation. As the Pew surveys on public perception of
FTA effects on jobs seem to confirm, American workers who
owe their jobs to rising exports are usually oblivious to their
dependence on foreign sales (in sharp contrast to workers who
lose their jobs to rising imports). Based on the increase in US
exports to Mexico, averaging $25 billion annually between
2009 and 2013, about 188,000 new US jobs were supported
each year by additional sales to Mexico. The figure is almost
A large portion of two-way trade
among the NAFTA economies represents
imported intermediates that raise
the competitiveness of US firms,
enabling them to improve their
export profile in world markets.
as large as the jobs lost, but the jobs gained in other sectors
pay better. On average, the export-related jobs pay 7 to 15
percent more than the lost import-competing jobs.20 The wage
differential, while positive, is only part of overall US gains
from trade with Mexico. In recent years, net US jobs lost on
account of two-way trade with Mexico have averaged about
15,000 annually (203,000 jobs displaced by imports minus
188,000 jobs supported by imports). Lawrence’s calculations,
cited earlier, suggest that gains to the US economy average
several hundred thousand dollars per net job lost.
Amidst the arithmetic of jobs lost and gained, it should
not be forgotten that a large portion of two-way trade among
the NAFTA economies represents imported intermediates
that raise the competitiveness of US firms, enabling them to
improve their export profile in world markets. In other words,
imports benefit not just US consumers but also US firms that
can acquire just the right intermediate components at the
right price.
The uneven impact of gains and losses from trade liberalization has been partially addressed by public policy at least
since the 1960s when the TAA was introduced. TAA offers
20. Richardson (2005) estimates that export jobs generally pay wages 10 to
11 percent higher and that US multinational firms pay 7 to 15 percent higher
than firms that are not globally engaged.
MAY 2014
assistance (e.g., extra unemployment insurance, training benefits, etc.) for workers who are displaced by imports.21 The share
of displaced workers certified as eligible for TAA is relatively
small compared to the total number of displaced workers in
the overall economy: In 2011, only 104,000 workers were
certified for TAA, out of 4.3 million workers displaced for all
reasons (Lawrence forthcoming 2014). Even though the costs
per TAA participant remain relatively low, namely $3,600 in
2011 and $6,500 in 2012, the program is frequently attacked
in Congress, especially by Republican members, who argue
that the United States should not support an “entitlement
program” aimed at sustaining and training workers who lose
their jobs to import competition.
A separate issue relates not to jobs lost or gained but
to the overall unemployment rate. Critics claim that larger
trade deficits add to the unemployment rate. In a hypothetical
economy where everything else is held equal (ceteris paribus,
in economists’ jargon), this is true. But historically everything
else is not held equal, and rising trade deficits are usually
associated with falling unemployment. Figure 2 charts the
inverse correspondence between the US unemployment rate,
expressed as a percent of the labor force, and the US global
trade deficit, expressed in billions of US dollars. As the figure
shows, almost without exception, when the trade deficit rises,
the unemployment rate falls. Over the past 30 years, periods
of high import growth in the United States have usually been
associated with tight labor markets and fast economic growth,
rather than weak labor markets and a slack economy.22
History is full of examples of a country at virtually full
employment yet running a trade deficit. The United States
enjoyed full employment in the late 1990s (unemployment
below 4 percent), despite a surge in imports led by Mexico.
The US economy reached close to full employment in 2007
despite rapidly rising imports from China and a large bilateral
and multilateral trade deficit. History is also full of examples
of a country with serious (if disguised) unemployment without
a trade deficit (the last two decades in Mexico and China).
Thus the connection claimed by NAFTA critics between
larger trade deficits and higher unemployment is seldom
observed in real life.23 But this is not to deny that an appro21. To receive TAA benefits, a group of workers must petition the US
Department of Labor and prove that international competition was the cause
of their job loss.
22. Edwards and Lawrence (2013) emphasize the general disjunction between
trade deficits and unemployment.
23. For examples of the misleading link between trade deficits and unemployment, see David E. Bonoir, “Obama’s Free-Trade Conundrum,” New York
Times, January 29, 2014, http://www.nytimes.com/2014/01/30/opinion/
obamas-free-trade-conundrum.html?_r=0 (accessed on March 3, 2014); and
Robert E. Scott, “NAFTA’s Legacy: Growing U.S. Trade Deficits Cost 682,900
7
NUMBER PB14-13
Figure 2
MAY 2014
US trade deficit and unemployment rate, 1992–2013
Unemployment rate, percent
monthly trade deficit, billions of current US dollars
12
80
Trade deficit
Unemployment rate
70
10
60
8
50
6
40
30
4
20
2
10
0
1992
0
1993
1994
1996
1997
1998
2000
2001
2002
2004
2005
2006
2008
2009
2010
2012
2013
Source: Federal Reserve Bank of St. Louis, http://research.stlouisfed.org/.
24. See Bergsten and Gagnon (2012). Other plausible candidates are tax
preferences that favor production for export markets and official support of
export finance through the Export-Import Bank.
against $19.50 in the United States.26 Taking advantage of the
wage difference, firms shift work to Mexico and build new
plants there. This puts pressure on US wages, directly through
layoffs and indirectly when firms threaten to close down in the
United States and open in Mexico, unless workers accept a
lower pay packet.
Exhibit A for wage criticism is the US auto industry.
Because this industry and its links with Mexico and Canada
are iconic symbols, we trace important developments since
NAFTA in box 1.
Most NAFTA critics acknowledge that the world is more
complex than suggested by a simple comparison of Mexican
and US wages. Many factors come into play when considering
whether competition from Mexico creates measurable downward pressure on US wages. High worker productivity, ready
access to needed inputs, reliable power, and an honest business
environment all offset low Mexican hourly wages as reasons to
produce in the United States.
Scholars have attempted to sort out the balance between
competition from low-wage countries, such as Mexico, and
other factors that determine wage levels in the United States. A
powerful analytical construct, the Stolper-Samuelson theorem,
demonstrates that if “other factors” can be ignored—most
25. Moreover, in the view of Theodore Moran (communication to the
authors), the argument that the United States should try to eliminate the trade
deficit via tougher trade negotiations or firmer market-access demands vis-à-vis
its trade partners so as to increase US employment reflects faulty analysis.
26. The cited figures are from the US Bureau of Labor Statistics. The Bank
of America reports a figure of $2.50 per hour for the Mexican manufacturing
wage, but possibly that figure excludes extensive fringe benefits.
priate policy measure that boosts exports at times of high US
unemployment, without cutting investment or government
spending, could both reduce the trade deficit and lower the
unemployment rate. The policy tool kit contains few such
measures, but a sharp realignment of exchange rates is one
candidate.24 Repealing NAFTA or any other trade agreement
is not a plausible answer to excessive trade deficits, for the
simple reason that US exports would surely drop as much as
US imports, if not more.25
U S M A N U FAC T U R I N G WAG E S
A powerful charge leveled by NAFTA critics is that trade with
Mexico has enabled US firms to hold back wage gains and
even cut wages. Their argument is straightforward. The current
average manufacturing wage is $4.50 per hour in Mexico,
Jobs,” December 17, 2013, http://www.epi.org/publication/nafta-legacygrowing-us-trade-deficits-cost-682900-jobs/ (accessed on March 13, 2014).
8
NUMBER PB14-13
Box 1
MAY 2014
Hard times in the US auto sector?
Table B.1 shows the total auto trade of the three NAFTA countries with the world in current US dollars. While US auto trade
clearly expanded over 20 years, the direction of trade remained relatively constant, with US imports roughly double US
exports in both 1993 and 2013. Of the three countries, Mexico saw the greatest growth by far in its global two-way auto
trade, expanding more than 11 times since 1993, in current dollars. By comparison, in current dollars, two-way auto trade
doubled in the case of Canada and nearly tripled for the United States. (Between 1993 and 2013, the US personal consumption expenditures [PCE] price index increased by 46 percent, so the nominal figures need to be deflated by that amount to
calculate real growth in auto trade.)
Table B.1
North American auto trade with the
world (billions of current US dollars)1
1993
2013
Imports
Exports
Imports
Exports
Canada
25
35
69
58
Mexico
2
7
32
70
85
40
249
118
United States
1. Includes all road vehicle trade and parts thereof. Between 1993 and
2013, the US personal consumption expenditures price index increased by
46 percent.
Source: World Bank, World Integrated Trade Solutions (WITS) database.
The labor picture in the auto industry (parts plus assembly) among the NAFTA partners changed dramatically, as shown
in table B.2. US employment of autoworkers fell by nearly a third between 1994 and 2013, while Canada experienced a
10 percent decline. By contrast, Mexico enjoyed a massive expansion of auto employment. Should the total decline of US
auto employment be laid at the doorstep of NAFTA? Probably not. According to the US Bureau of Economic Analysis,1 total
value added by the vehicle manufacturing industry (parts and assembly) was slightly higher in 2012 than in 1993, after
accounting for inflation. Correspondingly, over this period, real value added per worker increased by 41 percent (since the
auto labor force dropped by 28 percent).
Table B.2
Persons employed in the auto
manufacturing sector, parts and
assembly (thousands)
United States
Mexico
Canada1
1994
1,168
122
128
2013
820
552
115
1. 2012 data is the latest available for Canada.
Sources: US Bureau of Labor Statistics, Statistics Canada, and Instituto
Nacional de Estadística y Geografía (INEGI).
The increase in labor efficiency, driven largely by advanced manufacturing technology, was not accompanied by a
comparable increase in real US wages, as shown in table B.3.2 Worker compensation (wages plus fringe benefits) increased
by about 19 percent between 1994 and 2012, only 1 percent a year. At $37 per hour in 2012, average compensation in the
auto industry was slightly higher than average compensation in all manufacturing, about $36 per hour in 2012.
(box continues)
1. US Bureau of Economic Analysis (BEA), GDP by Industry data, http://www.bea.gov/industry/gdpbyind_data.htm.
2. Adjusted for inflation by the chain-type PCE price index of the US BEA. It is important to note that the apparent Canadian wage increases were largely driven by
appreciation of the Canadian dollar.
9
NUMBER PB14-13
Box 1
MAY 2014
Hard times in the US auto sector? (continued)
Table B.3
Real hourly compensation cost per
hour for motor vehicles, trailers, and
semi-trailers (2012 US dollars per hour)1
Country
1994
2012
Canada2
24.45
36.59
Mexico
United States
4.84
7.79
31.54
37.38
1. Compensation costs include direct pay, social insurance expenditures, and
labor-related taxes. The figures are adjusted for inflation by the chain-type
personal consumption expenditures (PCE) price index of the US Bureau of
Economic Analysis.
2. Canadian 2012 data for ISIC 29 not available. Figures estimated using the
Canadian compensation cost for all manufacturers.
Note: For 2012, the industry is defined as ISIC 29 and for 1994 it is defined as
SIC 371. Includes the manufacture of motor vehicles for passengers or freight,
parts and accessories, and trailers.
Sources: US Bureau of Labor Statistics, International Labor Comparisons, 2013,
http://www.bls.gov/fls/ichccindustry.htm#29/; and Hourly Compensation
Costs for Production Workers in Manufacturing, 2006, http://www.bls.gov/fls/
flshcindsic.htm.
Data from the US Bureau of Labor Statistics shows that unit labor costs fell by some 40 percent in vehicle assembly and
parts manufacturing over the past two decades.3 NAFTA critics claim that the trade pact accounts for a large part of the
disparity between productivity gains and wage gains. But another reason for this disparity is that the density of union
membership in auto manufacturing fell over time. Table B.4 shows the clear downward trend since the 1980s, but it does not
appear that the trend accelerated following the signing of NAFTA. The shift of US industry to the southern states was a major
factor in declining union density, since those states are generally less friendly towards unions. In 1984, the southeastern
states were responsible for just 11 percent of US automobile and parts production, a share that reached 26 percent in 2011.4
Yet a third reason for the slow pace of wage growth in the auto industry was the bankruptcy, or near bankruptcy, of many
auto firms in the Great Recession of 2008–09.
Table B.4
Union density in US auto manufacturing1
Percent of workers
unionized
1983
1994
2004
2013
58.8
44.4
29.4
18.2
1. Includes parts and assembly.
Source: Union Membership and Coverage Database from the Current Population
Survey, http://www.unionstats.com.
3. US Bureau of Labor Statistics, Labor Productivity and Related Data, Division of Industry Productivity Studies, March 2014, http://www.bls.gov/lpc/#tables.
4. See US Bureau of Economic Analysis (BEA), Regional Data. Figures based on SIC code 95000 for 1984 and NAICS code 95000 for 2011. As defined by the BEA,
the southeast region of the United States includes: Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee,
Virginia, and West Virginia.
importantly, technology differences between countries—and if
wage changes ripple across the labor market just like interest
rate changes ripple across the bond market, then import competition from low-wage countries will depress average US wage
levels (see Stolper and Samuelson 1941). In reality neither
10
assumption truly holds. The empirical question is whether the
assumptions underlying Stolper-Samuelson are close enough to
reality to generate the predicted outcome.
What does empirical research show? A recent study by
Autor et al. (2013) found that increased US imports from
NUMBER PB14-13
China between 1992 and 2007 did exert a modest negative
effect on US wages in manufacturing, reducing average earnings
in affected industries by roughly 3 percent from the base-year
level.27 By contrast, increased imports from Mexico and Central
American countries had no significant effect on US wages in
the manufacturing sector. This is true even though the United
States has engaged in substantial trade liberalization with its
Mexican and Central American trading partners and despite the
fact that in 2007 imports from Mexico and Central America
($233 billion) totaled over two-thirds that of imports from
China ($340 billion). Possibly the main reason the wage impact
between Chinese and Mexican imports differs is that US trade
with Mexico is roughly balanced and has a large intraindustry
component (e.g., autos and parts shipped in both directions),
whereas US trade with China is highly unbalanced and entails
Increased imports from Mexico and Central
American countries had no significant effect
on US wages in the manufacturing sector.
very large US imports of consumer goods in exchange for much
smaller US exports of capital goods. Because of these features,
US imports from Mexico compel considerably less job churn
between industrial sectors than US imports from China, and
this could account for the difference in estimated wage impact.
Like Autor et al. (2013), McLaren and Hakobyan (2010)
reported that, as a result of NAFTA, local US manufacturing
wages were not reduced, nor was there an industry-wide depression of wages. A separate study by Autor, Dorn, and Hanson
(2013) found that while the number of manufacturing jobs fell
due to increased imports, average manufacturing wages within
“commuting zones”—defined as metropolitan areas and their
surrounding localities—were not significantly affected by rising
imports. This finding will surprise observers who assume that
labor markets are fungible, akin to bond markets. Reviewing
the overall pattern of wages, Edwards and Lawrence (2013)
found that while workers at the lower end of the pay scale did
not experience much improvement over the past decade, they
did not fare much worse than the middle class, suggesting that
globalization has not significantly affected the distribution of
wage outcomes within the middle 80 percent of the American
economy. Whether globalization makes a major or minor
contribution to the good fortunes of the “top 1 percent” is
another question, one that we do not explore.28
27. In other words, if base-year manufacturing wages in an affected industry
were $45,000 in 1992, manufactured imports from China would reduce the
base year by $1,350.
28. Lawrence (2008) argues that the growing income share of the “super rich”
MAY 2014
NAFTA critics place far more emphasis on case examples
(“anecdotes”) than on the statistical analysis just reported. So
do the public and many politicians. Instances can certainly be
cited where import competition from Mexico led to wage cuts
in US plants or where the threat of moving a factory to Mexico
was used for leverage in wage negotiations. A study conducted
by Cornell University for the North American Commission
for Labor Cooperation found evidence that, between January
1993 and December 1995, over 50 percent of companies in the
United States threatened to close all or part of their production
plants in response to union activity or organizing campaigns of
workers (Bronfenbrenner 1996). Specifically, companies made
direct threats to relocate to Mexico in more than 10 percent of
the cases, while other cases involved implicit threats, such as
“given NAFTA we may need to reconsider our options” (p. 2).
In one case, ITT Automotive in Michigan underlined the threat
by setting up a dozen tractor-trailers full of production equipment from a closed site, plastered with signs reading “Mexico
Transfer Job.” In another case, a company handed out statistics to its workers on the differential between average wages
of Mexican and US autoworkers. Bronfenbrenner (1996 p. 3)
concluded that “NAFTA created a climate that has emboldened
employers to more aggressively threaten to close, or actually
close their plants to avoid unionization.”
But the anecdotes and the Bronfenbrenner (1996) survey
simply do not support the conventional wisdom that competition from Mexico has been a major force in suppressing the
growth of average US wages over the past two decades. Empirical
evidence in the cited studies indicates that increased imports do
decrease the overall number of manufacturing jobs. However,
increased imports of manufactures exert, at most, modest and
highly localized downward pressure on wages. Import competition has not so far created measurable downward pressure on
average wages, nor even on the wages of those who keep their
jobs in the manufacturing sector.
is largely driven by forces other than international trade. Those forces include
technology that amplifies the market scope of top entertainers (like Stephen
Colbert) and computer geniuses (like Eric Schmidt), financial deregulation,
plus changes in US corporate governance and rising share prices. Haskel et al.
(2012) find “suggestive evidence” that globalization has contributed to rising
earnings of superstars (defined as the small group of highly skilled, highly
compensated workers), but likely through globalization channels other than
merchandise trade, such as improved tradability of services and larger markets
abroad (p. 136). Kaplan and Rauh (2007) argue that trade is a poor explanation of increasing inequality, since the shift towards top earners extends well
beyond the sectors that produce tradable goods and services.
11
NUMBER PB14-13
U S AG R I C U LT U R A L E X P O R T S A N D M E X I C A N
I M M I G R AT I O N
Within a decade after the launch, NAFTA critics claimed that
US agricultural exports to Mexico had driven peasant farmers
from the land who then continued straight north to cross the US
border as illegal immigrants.29 The alleged damage was said to
be especially severe for cultivators of corn (maize), a staple crop
in hilly and arid agricultural districts, supposedly undercut by
huge corn exports from US agribusiness.30 To be sure, NAFTA
required some liberalization of Mexican corn imports. But
Mexico, of its own accord and in an effort to lower food prices
and control inflation, unilaterally accelerated liberalization by
allowing tariff-free imports of corn almost every year since 1994.
But the cause-and-effect story that labels US corn exports as the
cause of illegal Mexican immigrants does not stand up.
First, US corn exported to Mexico (the yellow variety) is
predominantly consumed by animals, whereas most corn grown
in Mexico (the white variety) is largely consumed by people
(tortillas and the like). Huge US exports of yellow corn have
enabled Mexicans to sharply increase the share of chicken and
beef in their daily diet. It has not replaced white corn.
Second, as in other emerging countries, the Mexican population is moving from the countryside to cities. Rural life in
most of Mexico is harsh, and incomes are barely 50 percent of
the urban average according to 2012 statistics from Mexico’s
National Institute of Statistics and Geography.31 However, in
the NAFTA era, the rate of rural-to-urban migration has actually decelerated. In the 20 years between 1970 and 1990, the
rural share of the Mexican population dropped 15 percentage
points, from 42 percent to 27 percent. In the 20 years between
1990 and 2010 (the NAFTA era), the decline was only 5
percentage points, from 27 percent to 22 percent.
Third, to maintain rural incomes, the Mexican government
has consistently supported the price of white corn with subsidies for farmers. In recent years, the average wholesale price of
white corn in major producing states ranged from $5.30 per
bushel in 2000 to $9.68 per bushel in 2013.32 One consequence
29. For an example, see Ellen R. Shaffer, “Immigration Is a NAFTA Problem.
This Is Not Big News,” Huffington Post, July 10, 2010, http://www.huffingtonpost.com/ellen-r-shaffer/immigration-is-a-nafta-pr_b_642484.html (accessed
on March 12, 2014).
30. For a detailed exploration of this corn saga, see Hufbauer and Schott
(2005), pp. 328–44.
31. Based on the 2012 National Survey of Mexican Household Income and
Expenditures, available at http://www.inegi.org.mx/est/contenidos/Proyectos/
Encuestas/Hogares/regulares/Enigh/Enigh2012/tradicional/default.aspx.
32. According to the Food and Agriculture Organization (FAO) Global
Information and Early Warning System (GIEWS) food price and data analysis
tool, http://www.fao.org/giews/pricetool/. The original prices were expressed
in US dollars per metric ton; these prices were converted to bushels using a
conversion factor of 39.37 bushels = 1 metric ton.
12
MAY 2014
is that the area under corn cultivation in Mexico has declined
only modestly in the NAFTA era, despite predictions that these
areas would be wiped out. In 1994, Canada, the United States,
and Mexico cultivated 1.0, 29.3, and 8.0 million hectares of
corn respectively. By 2013, these figures reached 1.5 million
hectares for Canada, 35.5 million for the United States, and 6.8
million for Mexico.33 Mexican cultivation decreased modestly,
but the big expansion of US corn cultivation reflects the ethanol
mandate—which in turn raised the global price of corn—not
exports to Mexico.
Fourth, there is little or no connection between the pace
of illegal immigration from Mexico and the level of US corn
exports to Mexico. Mexicans migrate to the United States to
earn a better living. But depending on the state of the US
economy, it can be easier or harder for an illegal immigrant to
land a job. Figure 3 shows the general correspondence between
the annual US unemployment rate and the number of border
apprehensions (the best proxy for the annual number of illegal
immigrants). No surprise: Higher unemployment discourages
illegal immigration because Mexicans are less likely to cross the
border if job opportunities are scarce. Tough border control has
an effect as well. Figure 3 also shows the annual tonnage of US
corn exports to Mexico (both yellow and white varieties). Visual
inspection fails to reveal a positive correspondence between
corn exports and apprehensions (if anything, the correspondence is negative). Even fancy econometrics cannot support the
claim that US corn exports drive Mexican migrants across the
Rio Grande. The corn-immigration story was a myth created
by US critics in their stretch to create an alliance between antiimmigrant forces and anti-NAFTA forces.
T H E B R O A D A S S O C I AT I O N B E T W E E N N A F TA
A N D I L L E G A L I M M I G R AT I O N
In the original NAFTA debate, President Carlos Salinas
famously framed NAFTA as a “choice between getting Mexican
tomatoes or tomato pickers,” while President Clinton predicted
that NAFTA would curb illegal immigration “because more
Mexicans would be able to support their families by staying at
home.”34
In their assessment, Hufbauer and Schott (1993) were
skeptical of both claims. In the short run, they argued, illegal
immigration would likely increase, both because of the huge
wage differential between the United States and Mexico and
because of the general movement of the Mexican population to
cities on the northern border (Tijuana, Juárez, Nuevo Laredo,
33. Figures from the US Department of Agriculture, Production, Supply and
Distribution database, http://apps.fas.usda.gov/psdonline/psdHome.aspx.
34. Quoted in Heyer (2012).
NUMBER PB14-13
Figure 3
MAY 2014
Illegal immigration, US unemployment, and US-Mexico corn trade
US unemployment rate, percent
thousands of apprehensions, thousands of metric tons
12
10,000
US border apprehensions from Mexico
US corn exports to Mexico
US unemployment rate
9,000
8,000
10
7,000
8
6,000
6
5,000
4,000
4
3,000
2,000
2
1,000
0
0
1994
1997
2000
2003
2006
2009
2012
Notes: Due to data availability, total apprehensions for the US southwest border are used as a proxy for Mexican apprehensions for 1994–98. 1999–2013 is
data for all apprehensions of Mexicans.
Sources: US Bureau of Labor Statistics; US Border Patrol Fiscal Year 2013 Statistics; USDA Foreign Agricultural Service, Global Agricultural Trade System
(GATS), 2014.
and others).35 In any event, immigration was essentially ignored
in the NAFTA text; the only exception was a limited number of
visas for business and professional migrants.
Taking a longer view, which Philip Martin (2005) does,
as the Mexican fertility rate falls and the population ages, the
number of young Mexicans who want to relocate in the United
States will diminish. This is the famous “demographic hump,”
first analyzed by Martin as a cause of immigration pressure.
In fact, as figure 3 shows, illegal immigration has noticeably
diminished since 2000, a combined outcome of three forces:
The demographic hump; higher US unemployment, especially
since 2007; and much stronger border controls.
Moreover, if the Mexican economy performs better in the
next two decades than in the past two (the subject of the next
section), the flow of illegal immigrants will diminish further.
To the extent that the Mexican economy flourishes as a result
of integration between the United States and Mexico, fostered
by NAFTA, the pact will serve as a positive force for higher
incomes and better living standards in Mexico and therefore
35. The portion of the Mexican population living in Mexican states that
border the United States has grown from about 16 percent in 1990 to 18
percent in 2013, according to data from the Instituto Nacional de Estadística
y Geografía (National Institute of Statistics and Geography).
diminished immigration to the United States. President
Clinton’s optimistic forecast will eventually prove right, but it
may take two generations for Mexican per capita incomes to
converge to half of the US level.
M E X I C A N G R O W T H I N T H E N A F TA E R A
Mexican growth in the NAFTA era has been disappointing. In
the wake of substantial economic reforms, Mexico should have
delivered a performance as good as Chile’s. It did not. Figure
4 compares real per capita GDP levels (adjusted for inflation),
between 1993 and 2013, for four relevant countries and country
groups: Mexico, Chile, the “ASEAN-4” (Indonesia, Malaysia,
the Philippines, and Thailand), and the “Andean-3” (Bolivia,
Ecuador, and Venezuela). Over the two-decade period, Mexican
real GDP per capita expanded 31 percent, which works out to
1.3 percent annually (compounded), whereas Chile expanded
90 percent, 3.1 percent annually. The ASEAN-4 expanded 75
percent, 2.7 percent annually, while the Andean-3 expanded 24
percent, only 1.0 percent annually.
Why did Mexico perform more poorly than Chile or the
ASEAN-4? Not because of NAFTA or lagging exports. Between
1993 and 2013, Mexican exports expanded 640 percent,
13
NUMBER PB14-13
Figure 4
MAY 2014
Real GDP per capita levels for select countries
US dollars (adjusted for inflation)
10,000
1993
9,000
2013
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
Mexico
Chile
ASEAN-4
Andean-3
Note: ASEAN-4 consists of Indonesia, Malaysia, the Philippines, and Thailand. Andean-3 consists of Bolivia, Ecuador, and Venezuela.
Source: ERS baseline dataset, US Department of Agriculture.
Chilean exports expanded 730 percent, and ASEAN-4 exports
expanded 420 percent.36 Instead, Mexico suffered from three
handicaps that were not nearly so burdensome in Chile and the
ASEAN-4. Foremost was organized mayhem stemming from
drug wars driven by the craving “made in the USA.” Drug
cartels have not only killed 70,000 people just since 2006,37
spreading fear across Mexico; they have also knocked GDP
growth down by around 1 percent annually. Other causes
of the lagging Mexican performance include weak primary
and secondary education; poor infrastructure (water, sewer,
gas, electricity, roads) in major urban areas, discouraging the
migration from farm to city; extensive corruption (compared
to Chile); persistent monopolization of key sectors (telecoms,
television, petroleum, electricity, cement); and sundry tax and
regulatory obstacles that stifle small business firms.
In fact, a McKinsey Global Institute report (2014) finds that
sectors of the Mexican economy oriented towards NAFTA—
primarily large firms employing 500 persons or more—enjoyed
36. Data from the International Monetary Fund (IMF), Direction of Trade
Statistics. Total exports for 2013 estimated based on the first three quarters.
The figures are in current dollars, not adjusted for inflation.
37. Nik Steinberg, “End Mexico’s Disastrous ‘War on Drugs’ Once and
For All,” Human Rights Watch, December 2, 2013, https://www.hrw.org/
news/2013/12/02/end-mexicos-disastrous-war-drugs-once-and-all (accessed
on March 5, 2014).
14
productivity growth of 5.8 percent annually between 1999 and
2009. The Mexican productivity problem is concentrated in
traditional small firms—employing 10 or fewer persons—which
have little connection to NAFTA. These firms account for 42
percent of the Mexican labor force, but their productivity actually declined between 1999 and 2009, dragging down the overall
growth of the Mexican economy.
As mediocre as Mexican GDP performance was for two
decades, it could have been worse. Look no further than the
Andean-3 to see the adverse impact—in per capita income
levels as well as growth—of populism, macroeconomic follies,
and deep state intervention. Conceivably, if the US Congress
had rejected NAFTA and refused to throw Mexico a financial
lifeline following the peso crisis of 1994, Mexican political and
economic policies might have taken a sharp left turn. Instead
of growing real per capita GDP at 1.3 percent annually, the
Mexican economy might have followed the trajectory of the
Andean-3, possibly shrinking per capita GDP, and the Mexican
political system might be rejecting new reforms rather than
tackling the problems of the state-owned petroleum company
Pemex and entrenched private monopolies.
NN Uu Mm Bb Ee Rr PP BB 11 44 -- T1 B3 D
Figure 5
m
M aA yY 2 0 1 4
NAFTA two-way merchandise trade
exports plus imports, billions of 2013 US dollars
1,400
Additional trade increase
1,200
Increase explained by GDP growth
Pre-FTA trade
1,000
800
600
400
200
0
Canada
Mexico
United States
FTA = free trade agreement
Note: Pre-FTA two-way trade is prior to the Canada-US FTA (1988) for Canada, and prior to NAFTA (1993) for the United States and Mexico. The middle part
of each bar is the increase in two-way trade explained by growth in real GDP up to the year 2013 for all countries, annualized based on the first three quarters
of 2013.
Source: International Monetary Fund, Direction of Trade Statistics, 2014.
W h at C U S F TA a n d NA F TA Ac h i e v e d
Figure 5 shows bar graphs for the North American two-way
merchandise trade of the United States, Canada, and Mexico,
respectively, expressed in 2013 dollars, adjusted for inflation.
The bottom portion of each bar (blue) shows the country’s
two-way trade prior to CUSFTA (1988) for Canada and prior
to NAFTA (1993) for the United States and Mexico.38 The next
segment of each bar (green) shows the amount of two-way trade
in 2013 that corresponds to North American GDP growth—in
other words, “business as usual” trade. The top segment of each
bar (light green) shows the country’s “extra” two-way trade. For
reference, table 1 presents much the same data underlying figure
5 but expressed in current dollars.39
Judging from these simple bars, “extra” US merchandise
trade is some $635 billion, about 55 percent of total North
American trade for the United States; $247 billion and 37
percent for Canada; and $345 billion and 63 percent for
Mexico. Of course CUSFTA and NAFTA cannot claim credit
for all the “extra” trade, but the agreements can claim credit
for a good portion.
CUSFTA and NAFTA have not exerted the same buoyant
impact on North American services trade as they have on
merchandise trade. Table 2 summarizes US services trade with
Canada, Mexico, and the world in 1993 and 2013, expressed
in current dollars. Imports and exports within North America
grew no faster, and sometimes slower, than with the world.
38. To be precise, the start date for Canada-US trade is 1988, and the start
date for US-Mexico and Canada-Mexico trade is 1993.
39. Table 1 is based on data from the US Bureau of Economic Analysis, while
figure 5 is based on data from the IMF Direction of Trade Statistics.
The Canada-US Free Trade Agreement (CUSFTA), signed in
1988, was the precursor to NAFTA, signed in December 1992,
and should be grouped with the trilateral pact when considering
achievements delivered by North American economic integration. CUSFTA and NAFTA were foremost trade and investment agreements, but of course they conveyed a larger message
of North America cooperation. Thus we start with “hard”
economic statistics and then move to “soft” political aspects.
Economic Payoffs
12
15
NUMBER PB14-13
Table 1
MAY 2014
US trade in goods (billions of current
US dollars)
1993
2013
Imports
Exports
Imports
Exports
Canada
113
101
338
302
Mexico
40
41
287
227
World
779
589
2,294
1,590
Source: US Bureau of Economic Analysis.
Table 2
US trade in services (billions of current
US dollars)
1993
Imports
Canada
Mexico
World
9
2013
Exports
17
Imports
30
Exports
64
8
11
17
29
124
186
428
660
Source: US Bureau of Economic Analysis.
While CUSFTA and NAFTA both contained services chapters,
and while investment in some service sectors was liberalized
(e.g., banking and retail trade in Mexico), regulatory barriers
to cross-border trade in services were not much reduced.
To this day, they remain high: One recent study by Centre
d’Etudes Prospectives et d’Informations Internationales, or
CEPII (Fontagné, Guillin, and Mitaritonna 2011), estimates
the average tariff-equivalent barriers to cross-border service
imports as follows: 24.2 percent for Canada, 46.8 percent for
Mexico, and 30.5 percent for the United States.40
Ample econometric evidence documents the substantial
payoff from expanded two-way trade in goods and services.
Through multiple channels, benefits flow both from larger
exports and larger imports.41 As a rough rule of thumb, for
advanced nations, like Canada and the United States, an agreement that promotes an additional $1 billion of two-way trade
increases GDP by $200 million. For an emerging country, like
Mexico, the payoff ratio is higher: An additional $1 billion
of two-way trade probably increases GDP by $500 million.42
Based on these rules of thumb, the United States is $127
billion richer each year thanks to “extra” trade growth, Canada
40. Based on the simple average of the ad valorem equivalents for communications, finance, insurance, other business services, and other services.
41. For a survey of the channels, see Bradford, Grieco, and Hufbauer (2005).
The channels include more efficient use of resources through the workings of
comparative advantage, higher average productivity of surviving firms through
“sifting and sorting,” and greater variety of industrial inputs and household
goods.
42. See Hufbauer, Schott, and Wong (2010), appendix A, table A.2.
16
is $50 billion richer, and Mexico is $170 billion richer. For
the United States, with a population of 320 million, the pure
economic payoff is almost $400 per person.
Appraising trade through a mercantilist lens, figure 6 shows
the percentage of growth in real US exports to Canada, Mexico,
and the rest of the world in the CUSFTA-NAFTA era. The start
point is indexed at 100 for each destination. While US exports
to Canada have grown almost as fast as US exports to the rest
of the world (which includes fast-growing Asia), US exports to
Mexico have grown much faster. From the American perspective, NAFTA must be credited with this mercantilist payoff.
CUSFTA and NAFTA probably had little impact on
inward foreign direct investment (FDI) to Canada and the
United States, because both countries were already open, with
settled commercial law and property rights. But for Mexico,
NAFTA turned the page from policies that said to foreign
investors “stay far away!” to policies that said “come on in!”
And because of its new access to US and Canadian markets,
Mexico became an attractive location for companies around
the world. In 1993, Mexico’s inward stock of FDI was just $52
billion, about 7 percent of GDP. By 2012, the stock reached
$315 billion, some 27 percent of GDP.
Political Payoffs
Possibly the largest payoff—for the United States more important than the economic benefits—was the creation of a new
foundation for US-Mexican relations through NAFTA. Unlike
Canada, Mexico has not been a US military ally in a long list
of wars, running from the First World War to the Afghanistan
conflict. In international conflicts, Mexico plays a neutral role.
Mexico was the only Latin American country to maintain ties
with Cuba for the entire duration of the Cold War. More to
the point, beginning with Mexico’s expropriation of Jersey
Standard and other foreign oil companies in 1938, the official
Mexican attitude towards foreign investment originating in the
United States was antagonistic. Generations of young Mexican
schoolchildren learned that the United States had stolen vast
swaths of Mexican territory by annexing the Texas Republic in
1845 and by seizing present day New Mexico, Colorado, Utah,
Arizona, Nevada, and California in the Mexican-American War
(which was concluded by the Treaty of Guadalupe Hidalgo in
1848). Bearing this history in mind, it is fair to characterize
US-Mexican relations prior to NAFTA as cool.
NAFTA dramatically improved the dynamic of official
and private relations. The Clinton administration spearheaded
bilateral and multilateral assistance to Mexico in the wake of
the 1994 peso crisis. As drug wars escalated in the 2000s, the
United States provided, and Mexico welcomed, intelligence
NUMBER PB14-13
Figure 6
MAY 2014
US real export growth since the pre-FTA level
percent increase, constant US dollars
220
200
180
160
140
120
100
To Canada (since 1988)
To Mexico (since 1994)
To ROW (since 1988)
FTA = free trade agreement; ROW = rest of world
Note: Pre-FTA for Canada is pre-CUSFTA; pre-FTA for Mexico is pre-NAFTA.
Source: International Monetary Fund, Direction of Trade Statistics, 2013.
assistance and military supplies. Cooperation was good on
issues ranging from agricultural inspection to climate change
to border inspections. Certainly the United States could have
done more to foster integration with Mexico.43 NAFTA did
not address the thorny problem of a path to citizenship for
11 million undocumented Mexicans living in the United
States, an issue which continues to flare up in congressional
debates. But it seems highly unlikely that bilateral relations
over the past 20 years would have been equally cordial without
NAFTA.
In addition, NAFTA gets some credit for Mexico’s transition from a one-party political system with extensive state
capitalism to a multiparty market-oriented system—but of
course most of the credit for these reforms goes to internal
Mexican forces. NAFTA can also claim some credit for the
rise in Mexico’s ratings in the Heritage Foundation’s Index of
Economic Freedom, from 63 in 1995 to 67 in 2013.44
While political payoffs cannot be quantified in economic
terms, over the course of the next 20 years they are likely to
prove more consequential than the economic payoffs already
realized from the NAFTA pact.
43. The late Robert Pastor was a forceful exponent of more energetic and
generous US policies towards Mexico. See Pastor (2011).
44. For details on the Index of Economic Freedom, see http://www.heritage.
org/index/.
17
NUMBER PB14-13
APPENDIX A
MAY 2014
Table A.1
T R A D E AG R E E M E N T S A N D T R A D E D E F I C I T S
A favorite attack line of trade skeptics is that US free trade
agreements inevitably worsen US trade deficits.45 “Uncle
Sucker” is their metaphor: US negotiators open wide the gates
for foreign imports but gain precious little access for American
exports. The predictable outcome, they reason, is larger US
trade deficits. According to their arithmetic, trade deficits
translate into lost US jobs. The main text spells out the normal
inverse relationship between trade deficits and unemployment.
In this appendix we examine the starting point in the critics’
chain of logic: the supposed strong connection between trade
agreements and trade deficits.
F O R E I G N T R A D E B A R R I E R S FA L L M O R E
At the launch of a free trade agreement (FTA), US trade barriers
are almost always lower than the barriers of a prospective partner
country. Why? Because, unlike many countries, the United
States has progressively trimmed its trade barriers for 70 years,
since the end of the Second World War. Lower US trade barriers
at the launch of the agreement were true of the Canada-US FTA
(CUSFTA) in 1989; NAFTA, which added Mexico in 1994;
the US-Chile FTA in 2004; the US-Australia FTA in 2005; the
US-Peru FTA in 2009; the US-Korea FTA in 2012; and others.
Since the goal of any FTA is to reduce the trade barriers of both
partners to zero, the US partner almost always has further to go.
In other words, the partner opens its markets to US
exports more than the US opens its markets to partner exports.
Using tariff data, table A.1 illustrates this basic and important
fact, both for existing FTAs and the prospective Trans-Pacific
Partnership (TPP) and the Transatlantic Trade and Investment
Partnership (TTIP). FTAs also lower nontariff barriers (NTBs,
for example, quotas and regulatory obstacles) in both partners.
Extensive research shows that, like tariffs, NTBs are generally
higher in the partner country when an FTA is launched. Again,
the partner has further to go. So, just looking at negotiated FTA
texts, it is “Uncle Smart,” not “Uncle Sucker.”
45. Ross Perot started this line of attack, claiming that NAFTA would create
a “great sucking sound.” Prominent among current skeptics are David Bonior
(former Congressman), Thea Lee (AFL-CIO), Clyde Prestowitz (Economic
Strategy Institute), Robert E. Scott (Economic Policy Institute), and Lori
Wallach (Public Citizen’s Global Trade Watch).
18
US and FTA partner average
tariffs at year of entry
(unweighted)
US applied
tariff
Partner
applied tariff
Existing FTAs
Canada (1989)
5.06
9.65
Mexico (1994)
4.32
12.36
Australia (2005)
3.11
5.10
Peru (2009)
2.98
8.57
Colombia (2011)
2.79
11.17
South Korea (2012)
2.79
10.08
Malaysia (2012)
3.4
6.5
Vietnam (2012)
3.4
9.5
Japan (2012)
3.4
4.6
3.4
5.5
TPP (2015?)
TTIP (2016?)
European Union (2012)
FTA = free trade agreement; TPP = Trans-Pacific Partnership;
TTIP = Transatlantic Trade and Investment Partnership
Note: Year of entry into force or closest earlier available year for US FTA
partners, latest available year for prospective FTA partners. Tariff data
is from 1993 for Mexico, 2004 for Australia, 2006 for Peru, and 2011 for
South Korea.
Sources: World Bank, World Integrated Trade Solutions (WITS) database;
WTO tariff profiles; UN Conference on Trade and Development Trade
Analysis and Information System (UNCTAD TRAINS) database.
U S T R A D E D E F I C I T S W I T H F TA PA R T N E R S A N D
OT H E R S
According to the skeptics, the United States should be accumulating a mountain of trade deficits with its 20 FTA partners,
starting with Israel in 1985, then Canada and Mexico in the
1990s, and moving along to Korea, Colombia, and Panama in
the 2010s. But this hasn’t happened. Figure A.1 tells the story.
Excluding fuel imports and exports from the picture (if US oil
imports did not come from Canada and Mexico, they would
come at higher cost from Venezuela, Nigeria, and elsewhere),
it is evident that US trade deficits with its FTA partners are
coasting along at $50 billion to $100 billion annually, while
US deficits are mounting with the rest of the world. It would be
wrong to promise that the TPP or the TTIP would “cure” US
trade deficits with those countries. But it’s false to claim that
NAFTA and other FTA pacts are the locomotive driving higher
US trade deficits over the last two decades.
NUMBER PB14-13
Figure A.1
MAY 2014
US nonfuel merchandise trade balance, 1985–2012
billions of 2009 US dollars
100
0
– 100
– 200
– 300
– 400
– 500
FTA partners
Others
– 600
Source: World Bank’s World Integrated Trade Solutions (WITS) database, https://wits.worldbank.org/.
G LO B A L E X P E R I E N C E : T R A D E AG R E E M E N T S
AND TRADE DEFICITS
If the FTA skeptics were right, greater coverage of a country’s
imports by trade agreements should foreshadow larger trade
deficits. Figure A.2 examines this hypothesis. The vertical axis
portrays each country’s 2012 trade surplus (positive) or trade
deficit (negative) as a percentage of its GDP.46 The horizontal
axis shows the percentage of the country’s imports that are
covered by trade agreements (leading to tariff preferences). If
more trade agreements meant larger trade deficits, the country
dots would drift downwards, left to right in the figure. They
don’t. As far as the eye can tell, trade deficits are symmetrical
around the horizontal axis in figure A.2: In other words, more
import coverage by trade agreements doesn’t foreshadow either
larger trade surpluses or larger trade deficits. As the arrow in the
graph indicates, the United States is a middling country when it
comes to both trade agreements and trade deficits. In 2012, US
46. The technical term for trade surpluses and deficits is “current account
balance,” a concept which covers trade in goods and services and other current
transactions such as remittances.
preferential tariffs covered less than 20 percent of US imports,
and the US trade deficit was about 3.4 percent of US GDP.
W E L L , W H AT E X P L A I N S T R A D E D E F I C I T S ?
Though a household budget analogy might seem simplistic, it’s
not far off the mark. When a household earns $100,000 and
spends $105,000 on goods and services, that household has a
deficit of $5,000. The deficit must be financed by a mortgage
loan, credit card debt, or a generous relative. (We’ll assume that
neither Willy Sutton nor Bernie Madoff heads the household.)
Likewise, when a nation earns $15.7 trillion and spends $16.2
trillion on goods and services (approximately the US case in
2012), the national trade deficit will be $500 billion. That
deficit must be financed by loans or investment from abroad.
Figure A.3 portrays this basic and fundamental story in bar
graphs. The annual US trade deficit closely matches, year by
year, the combined deficiency in US net national savings (in
other words, net national borrowing). The combined deficiency
is the sum of household financial savings (or deficits), government deficits, and business savings (the difference between
company profits and company investments). Adding these three
19
NUMBER PB14-13
MAY 2014
Figure A.2 Trade balances and preferential imports, 2012
current account, surplus or deficit, percent of GDP
50
40
30
20
10
0
–10
–20
–40
–50
United States
–30
–60
percentage of imports entering on a preferential basis
Note: Preferential basis refers to the percentage of imports entering at less than half of the most favored nation rate. This corresponds closely to trade
within free trade agreements.
Sources: World Bank’s World Integrated Trade Solutions (WITS) database, https://wits.worldbank.org/; authors’ calculation.
components gives net national borrowing—in other words,
negative national savings. When net national borrowing goes
up, so does the trade deficit—because the borrowed money is
spent on foreign goods and services (or, to put the relationship
another way, because the United States spends more than it
earns, it must borrow from abroad).
If the United States wants to reduce its trade deficit, it
must reduce its net national borrowing. Many policies can help.
Government deficits can be cut, household financial savings can
be encouraged, foreign central banks can be asked to appreciate
their currencies and buy fewer US Treasury bonds, and the US
20
Export-Import Bank can enlarge its lending to foreign buyers
of US exports. In addition, as Bergsten and Gagnon (2012)
have urged, in periods of high US unemployment, the United
States might reduce its trade deficit by policies that realigned the
exchange rate—in plain English, a cheaper dollar relative to the
euro, the yuan, or the yen. But one policy that makes no sense,
and will do little or nothing to reduce the trade deficit, is to block
new FTAs—such as TPP and TTIP. Instead, these agreements
should be pursued vigorously for what they can bring: higher
productivity and better living standards for all Americans. That’s
what NAFTA promised and that’s what NAFTA delivered.
NUMBER PB14-13
Figure A.3
MAY 2014
US trade deficit and net national savings
billions of current US dollars
0
–100
–200
–300
–400
–500
–600
–700
Net national savings
–800
Trade balance
–900
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Note: Negative savings values indicate net borrowing. The trade balance reflects the current account balance, exlcuding investment income.
Source: Federal Reserve Bank of St. Louis, http://research.stlouisfed.org/.
REFERENCES
Autor, David H., David Dorn, and Gordon H. Hanson. 2013. The
China Syndrome: Local Labor Market Effects of Import Competition in
the United States. NBER Working Paper No. 18054 (May). Cambridge,
MA: National Bureau of Economic Research.
Autor, David H., David Dorn, Gordon H. Hanson, and Jae Song.
2013. Trade Adjustment: Worker Level Evidence. NBER Working Paper
No. 19226 (July). Cambridge, MA: National Bureau of Economic
Research.
Bergsten, C. Fred, and Joseph E. Gagnon. 2012. Currency Manipulation,
the US Economy, and the Global Economic Order. Policy Brief 12-15
(December). Washington: Peterson Institute for International
Economics.
Bernard, Andrew B., J. Bradford Jensen, and Peter Schott. 2003. Falling
Trade Costs, Heterogeneous Firms and Industry Dynamics. IFS Working
Paper 03/10. London: Institute for Fiscal Studies.
Bradford, Scott C., Paul L. E. Grieco, and Gary Clyde Hufbauer. 2005.
The Payoff to America from Global Integration. In The United States
and the World Economy: Foreign Economic Policy for the Next Decade,
ed. C. Fred Bergsten. Washington: Peterson Institute for International
Economics.
Bronfenbrenner, Kate. 1996. The Effects of Plant Closing or Threat
of Plant Closing on the Right of Workers to Organize. Final Report.
Submitted to the Labor Secretariat of the North American Commission
for Labor Cooperation (September 30). New York: Cornell University.
Available at http://digitalcommons.ilr.cornell.edu/cgi/viewcontent.
cgi?article=1000&context=intl (accessed April 17, 2014).
Edwards, Lawrence James, and Robert Z. Lawrence. 2013. Rising Tide:
Is Growth in Emerging Economies Good for the United States? Washington:
Peterson Institute for International Economics.
Fontagné, Lionel, Amélie Guillin, and Cristina Mitaritonna. 2011.
Estimations of Tariff Equivalents for the Services Sectors. No. 2011-24
(December). Paris: Centre d’Etudes Prospectives et d’Informations
Internationales (CEPII). Available at http://www.cepii.com/PDF_
PUB/wp/2011/wp2011-24.pdf (accessed April 17, 2014).
Haskel, Jonathan, Robert Z. Lawrence, Edward E. Leamer, and
Matthew J. Slaughter. 2012. Globalization and U.S. Wages: Modifying
Classic Theory to Explain Recent Facts. Journal of Economic Perspectives
26, no. 2 (Spring): 119–40. Available at http://www.hks.harvard.edu/
fs/rlawrence/JournalEconPerspectives.pdf (accessed April 17, 2014).
Heyer, Kristin E. 2012. Kinship Across Borders: A Christian Ethic of
Immigration. Washington: Georgetown University Press.
21
NUMBER PB14-13
MAY 2014
Hufbauer, Gary Clyde, and Jeffrey J. Schott. 1993. NAFTA: An
Assessment, Revised Edition. Washington: Institute for International
Economics.
McLaren, John, and Shushanik Hakobyan. 2010. Looking for Local
Labor Market Effects of NAFTA. NBER Working Paper 16535.
Cambridge, MA: National Bureau of Economic Research.
Hufbauer, Gary Clyde, and Jeffrey J. Schott. 2005. NAFTA Revisited:
Achievements and Challenges. Washington: Peterson Institute for
International Economics.
Moran, Theodore H. 2011. FDI and Development: Launching a
Second Generation of Policy Research: Avoiding the Mistakes of the
First, Reevaluating Policies for Developed and Developing Countries.
Washington: Peterson Institute for International Economics.
Hufbauer, Gary Clyde, Jeffrey J. Schott, and Woan Foong Wong.
2010. Figuring Out the Doha Round. Policy Analyses in International
Economics 91. Washington: Peterson Institute for International
Economics.
Kaplan, Steven N., and Joshua Rauh. 2007. Wall Street and Main Street:
What Contributes to the Rise in the Highest Incomes. NBER Working
Paper 13270. Cambridge, MA: National Bureau of Economic Research.
Available at http://www.nber.org/papers/w13270.pdf (accessed April
17, 2014).
Koopman, Robert, William Powers, Zhi Wang, and Shang-Jin Wei.
2010. Give Credit Where Credit Is Due: Tracing Value Added in Global
Production Chains. NBER Working Paper 16426. Cambridge, MA:
National Bureau of Economic Research. Available at http://www.bea.
gov/about/pdf/NBER%20working%20paper_1.pdf (accessed April
17, 2014).
Lawrence, Robert Z. 2008. Blue-Collar Blues: Is Trade to Blame for Rising
US Income Inequality? Policy Analyses in International Economics 85.
Washington: Peterson Institute for International Economics.
Lawrence, Robert Z. 2014 (forthcoming). Adjustment Challenges for
US Workers. In Bridging the Pacific: Toward Free Trade and Investment
Between China and the United States, eds. C. Fred Bergsten, Gary
Clyde Hufbauer, and Sean Miner. Washington: Peterson Institute for
International Economics.
Martin, Philip. 2005. Mexico-US Migration. In NAFTA Revisited:
Achievements and Challenges, eds. Gary Clyde Hufbauer and Jeffrey J.
Schott. Washington: Peterson Institute for International Economics.
McKinsey Global Institute. 2014. A tale of two Mexicos: Growth and
prosperity in a two-speed economy. Report prepared by Eduardo Bolio,
Jaana Remes, Tomás Lajous, James Manyika, Eugenia Ramirez, and
Morten Rossé (March). Mexico City: McKinsey & Company. Available
at http://www.mckinsey.com/insights/americas/a_tale_of_two_mexicos (accessed April 17, 2014).
Pastor, Robert. 2011. The North American Idea: A Vision of a Continental
Future. New York: Council on Foreign Relations.
Public Citizen’s Global Trade Watch. 2014. NAFTA’s 20-Year Legacy
and the Fate of the Trans-Pacific Partnership. Washington: Public Citizen.
Available at http://www.citizen.org/documents/NAFTA-at-20.pdf (accessed April 17, 2014).
Richardson, David J. 2005. Uneven Gains and Unbalanced Burdens?
Three Decades of American Globalization. In The United States and the
World Economy: Foreign Policy for the Next Decade, ed. C. Fred Bergsten.
Washington: Institute for International Economics.
Scott, Robert E. 2011. Heading South: U.S.-Mexico trade and job displacement after NAFTA. EPI Briefing Paper #308 (May 3). Washington:
Economic Policy Institute. Available at http://www.epi.org/publication/heading_south_u-s-mexico_trade_and_job_displacement_after_
nafta1/.
Scott, Robert E. 2013. No Jobs from Trade Pacts: The Trans-Pacific
Partnership Could Be Much Worse than the Over-Hyped Korea Deal. Issue
Brief #369 (July 18). Washington: Economic Policy Institute. Available
at
http://www.epi.org/publication/trade-pacts-korus-trans-pacificpartnership/ (accessed April 17, 2014).
Scott, Robert E. 2014. White House Wrong on Fast Track: Trade Deals
Cost Jobs, Depress Wages. Policy Memorandum #202 (February 5).
Washington: Economic Policy Institute. Available at http://www.epi.
org/publication/white-house-wrong-fast-track-massive-trade/ (accessed
April 17, 2014).
Stolper, W. F., and Samuelson, Paul A. 1941. Protection and Real
Wages. Review of Economic Studies 9, no. 1: 58–73.
This publication has been subjected to a prepublication peer review intended to ensure analytical quality. The views expressed
are those of the authors. This publication is part of the overall program of the Peterson Institute for International Economics,
as endorsed by its Board of Directors, but it does not necessarily reflect the views of individual members of the Board or of
the Institute’s staff or management. The Peterson Institute for International Economics is a private, nonprofit institution for
the rigorous, open, and intellectually honest study and discussion of international economic policy. Its purpose is to identify
and analyze important issues to making globalization beneficial and sustainable for the people of the United States and the
world and then to develop and communicate practical new approaches for dealing with them. Its work is made possible by
financial support from a highly diverse group of philanthropic foundations, private corporations, and interested individuals,
as well as by income on its capital fund. For a list of Institute supporters, please see www.piie.com/supporters.cfm.
22